Torsion Box Structural Design

[ChrisA Edwards]

I've made one small Torsion box in the past and am about to make another.

So I thought I'd ask the experts here.

This one will be 7' x 4'. It will support a hobbiest CNC (OneFinity) with has a footprint of 62" x 45". Weight of the gantry and spoil board, I guess, will be about the 150lbs.

The gantry and spoil board will also sit on an aluminum extrusion frame that is about 3" high.

My plan is to have this Torsion top sit on a cabinet with draws and cabinets, so the top will have support in the central areas below, not just the ends.

I plan on using 3/4" material for the top and bottom, most likely MDF for both.

So two basic questions:

1) I plan on using 3/4" material for the ribs. Is plywood or MDF better. I'm not worried about the cost difference.

2) Is there an internal rib height where the ribs are less likely to bow, i.e. 2", 3" etc.. With the aluminum chassis, that the gantry sits on, I'm thinking a supported Torsion box doesn't need super high ribs.

I also plan to fill the pockets inside the torsion box with some sound deadening material, looking for recommendations for this, or is this overkill?

I've thought about 2" Pink solid foam, regular fiber glass insulation, spray in expanding foam.

I will build an enclosure over the top of the gantry, so I'm trying to get this to where having a CNC spindle working, in the background, isn't an annoying nuisance.

Thanks

[Keegan Shields]

Something to keep in mind - a torsion box gets its strength from the height of the ribs (distance between the 2 skins) and the bond of the ribs to the skin.

https://en.wikipedia.org/wiki/Torsion_box

And if I remember correctly, the relationship of the rib height to strength is not linear - meaning a 2" rib height is more than 2x as strong as box with 1" rib height.

Here's a good SMC thread: https://sawmillcreek.org/showthread.php?131479-Torsion-box-height-vs-strength

Hope that helps,
Keegan

[Richard Coers]

I think you are missing the advantages of the torsion box if you use 3/4 MDF for the skins and ribs. You could park a truck on what you are proposing. Also MDF edge grain gluing is not that strong. You are gluing to compressed sawdust. You won't change the sound travel with all the solid ribs connecting the skins. You need a complete barrier to limit sound. Sound limiting is about reducing vibration. You are doing that with the ribs. Finally I see no need to make a torsion box if it is completely supported with a cabinet under the whole top.

[Mike Henderson]

Keegan has some good advice. From my experience, the problem you run into is the weight of the box. You do not need 3/4 inch material for the ribs - much thinner material will work fine if you glue it well to the top and bottom. 1/4 inch plywood would work, and 1/8 might also work and keep the weight down.

The thickness of the top and bottom is mainly determined by the space between the ribs. If you have close spacing of your ribs, you can go with a thinner top and bottom. The limit is how much deflection there is in the top (and bottom) between ribs.

I wouldn't use MDF just because of the weight.

Mike

Note that in some doors, they use a grid of fairly thin cardboard for the ribs and the surface material is quite thin.

[Jamie Buxton]

+1 on the thickness of your materials. I have a assembly table which is a torsion box I put on sawhorses. The torsion box is 42x72x4.5”. It weighs 26 pounds. If i put the sawhorses under the ends, and sit in the middle, it sags less than 1/32”. The spacer grid is an egg crate made with 1/8” plywood on 6” centers. The skins are quarter inch plywood.

[John TenEyck]

No need to guess. Calculate: https://steve68steve68.blogspot.com/2008/02/torsion-box.html

And I agree with Richard; if the machine's feet sit on solid support it won't flex.

John

[Kevin Jenness]

4" thick is good for that size table. I worked for years on a 4" x 48" x 96" torsion box skinned with plastic laminate and built similar to Jamie's with1/4" lauan skins and 5/16" basswood grid core @5"o.c. Very stiff and light.

You could also use foam for the core. I suspect that would be deader (more dead?) acoustically than a grid.

[Christopher Wellington]

Something to keep in mind - a torsion box gets its strength from the height of the ribs (distance between the 2 skins) and the bond of the ribs to the skin.

https://en.wikipedia.org/wiki/Torsion_box

And if I remember correctly, the relationship of the rib height to strength is not linear - meaning a 2" rib height is more than 2x as strong as box with 1" rib height.
Keegan, you're correct about the height of the rib being more important than the width. The stiffness of a beam (how little it deflects under load) is proportional to its width and to the cube of its height.

The term of interest in the beam deflection equations is a geometric value referred to as "area moment of inertia." For a rectangle, that value is equal to (b*h3)/12 where b is the base (or width) and h is the height. Double the width of the beam and you only double its stiffness. Double the height and you increase its stiffness by a factor of 8!

If you're concerned about weight and stiffness (isn't that why the torsion box is so good?) use tall and narrow ribs.

https://www.engineeringtoolbox.com/area-moment-inertia-d_1328.html

[Malcolm McLeod]

Keegan, you're correct about the height of the rib being more important than the width. ...

^+1
An airplane wing is essentially a torsion box, and its skinned in thin aluminum sheet. The internal structure keeps the wing skin loaded almost entirely in tension - where it is quite strong, but if loaded in compression will buckle. (You can use a strip of tinfoil and tow your milk jug across the table, but just try to push it with that same strip.:eek:)

I think you'll find the same principle applies to a torsion box 'table' >> with properly sized ribs, the top and bottom skins could be <1/8" thick (aka Mr. Buxton's 26lb top?).

[Wayne Cannon]

Re: Sound Deadening
Thermal insulation and sound deadening qualities are almost opposite. Insulation relies on trapping air, but air does not attenuate sound. Fiberglass batting (and shrubbery) do very little to reduce sound. Dense foam would be better, but rubber even better.

It's the flexing of the cell walls that attenuate sound by converting it into heat. Density also helps by reducing amplitude. A solid rubber mat, a dense rubber floor mat (e.g., for horse stalls), or sheet-rock panel under your CNC will help deaden conducted sound. Dense enclosure walls do the same for radiated sound. Avoid fasteners (screws, nails, battens) that bypass the absorbing material.

In sound studios, we used two layers of sheet-rock separated by a dense foam rubber sheet and separated from the studs/joists by flexible clips, and double alternating-stud walls.

[John Stankus]

Design info from APA the Engineered wood Association
https://law.resource.org/pub/us/code/bsc.ca.gov/sibr/org.apawood.U813.pdf

The appendix has a simplified method that might be easier to deal with.